Sam Everington - Engine by Starling: Why banks are renting Starling's core instead of building their own.

In 2016, Sam Everington joined Starling Bank as one of its first twenty employees. Starling did not have a banking licence yet. It had no product, no brand, and no customers. What it had was a room full of engineers who believed a bank could be built from the customer support desk backwards, rather than from a requirements document forwards.

Ten years and one banking licence later, Starling Bank has 4.5 million UK customers, half a million business accounts, and a 9% share of UK business banking. I met Sam and his colleague Mark Bernhardi, who runs Engine's Australia and New Zealand business, at the World Credit Union Conference in Sydney. What they described wasn't a fintech growth story. It was an answer to a question most bank boards are quietly avoiding: can you fix a core banking problem without a decade and hundreds of millions of dollars.

Engine by Starling: the neobank that became a core banking vendor

The easy version of this story is that Starling did well, and one of its early engineers is now senior. That's true, and it's the least interesting part of it.

Four years ago, Starling split its technology into a separate company, Engine by Starling, and started selling the core banking platform it built for itself to other banks. AMP Bank went live on Engine in under twelve months. SBS Bank in New Zealand signed a ten-year deal with Engine in February, its first mutual client. Salt Bank in Romania and Tangerine in Canada run on the same platform. That is not a digital bank licensing its brand. That is a neobank turning its core technology into a product category.

Why the neobank speed problem was always a core banking problem

Originally it was thought the neobank threat to incumbents was about interfaces: better apps, friendlier onboarding, cheaper fees. That was Fintech 1.0, the low hanging fuit. The real cost incumbents carried sat underneath, in core banking systems nobody outside a technology division ever sees, and in the capital required to replace them.

Sam puts the figure at hundreds of millions of pounds to build what Engine now licenses. Mark Bernhardi, who spent years selling core banking transformation at nCino, described the mechanism plainly: every bank on Engine runs the same software Starling runs, updated 40 to 60 times a day. A mutual bank in New Zealand and a digital bank in Romania are, in a real sense, running the same bank.

The commercial model is just as important as the technology. Engine is sold as a managed service rather than a licensed stack a bank has to run itself, which means the heaviest capital outlay only arrives once the platform is delivering usage, not before a single customer has signed up. That is a different risk profile to the traditional core banking RFP, where the cheque is written years before any benefit shows up, and it is the detail that gets a business case past a CEO who has already been burned by one transformation program.

A decade of watching core banking eat bank strategy

This is where my own decade in fintech recruitment lines up with theirs. I've watched two waves in Australian banking since 2016. The first was the restricted authorised deposit-taking institution licence wave of 2018 and 2019, when a run of digital challengers launched believing they could compete with the Big Four who have 85% market share. Almost none of them survive as standalone brands today.

The second wave, the one underway now, is quieter and involves no new logos at all. It is existing institutions, mostly mutuals and second-tier banks, buying the core banking technology stack a neobank would have built, rather than the neobank itself.

The first wave mostly failed for the reason Sam gave for Engine's existence: the big four were never going to lose on technology, because the profits that followed the GFC let them keep reinvesting in it while smaller players tried to catch up. A new licence without a distribution advantage or a technology advantage is just a smaller, less capitalised version of an incumbent. What changes the equation for a mutual bank or a credit union isn't a new licence. It's access to a core banking platform someone else already spent a decade paying for, priced as a service rather than a build.

SBS Bank is a member-owned mutual whose balance sheet wouldn't stretch to a fraction of what a major bank spends on technology in a single year. If Engine's model works there, the constraint that has kept Australia and New Zealand's smaller customer-owned banks running on ageing cores is no longer there.

The obvious objection is risk: touching the core has long been the accepted career killer for a banking CEO, because the value of a core migration typically lands in year four or five, well past the tenure of the executive who approved it and took on all the downside. Engine's answer is to compress that timeline, moving the growth benefit into year one and the full payback into year three, which is a genuinely different risk profile for a board to sign off on, including under stress testing.

What Engine by Starling means for digital banks in Australia and New Zealand

The next decade will be won by whoever can absorb change fastest without breaking governance, and that capability is now something a bank can buy off the shelf rather than something it has to build over a decade the way Starling did.

There is already a second order effect visible in the market, and it has nothing to do with cost. Bank leadership teams that get a modern core banking platform live quickly start talking about product with more conviction, not because pricing changed, but because they can finally ship something differentiated instead of a marginally cheaper version of the same term deposit. That shift in ambition, from defending margin to building product, is arguably a bigger prize for a bank's board than the capex saving, because it changes what the executive team believes is possible to build in the first place.

For Australia and New Zealand's customer owned banking sector specifically, that's a genuine opportunity to compete for customers. A mutual bank was never going to out invest CBA on technology. What it can do is leverage the same technology platform that has the major banks worried.

The talent question Engine by Starling raises for every bank board

Ten years ago, Sam Everington sat in a room with nineteen other people and no banking licence, building the systems that would eventually serve 4.5 million customers. What he described in Sydney wasn't the end of that story. It was that same build, packaged, priced, and now available to any bank willing to buy speed instead of spending a decade building it.

Buying the platform is the easy half of this decision. The harder half is who runs it once it lands. A core banking migration compressed from five years into one still needs an operator who has built and shipped inside exactly that kind of ambiguity, not a career banker who has only ever managed a core that was already stable. That's a different hire than the one most banks have on their bench today.

Racing to fix a legacy core and need the operator who can lead it? Talk to us about your search.